New method of tumor detection using K-means classifier and thresholding process
The diagnosis of cerebral tumors is a process that identifies the cause and the existence of a tumor in the human brain. The diagnostic process may seem long and discouraging but it is necessary and important. To reduce waiting time of patients; this time which is venomous and very exhausting, the researchers opted for automating certain stages of diagnosis, today thankfully the detection and location of the tumor can be achieved by MRI using a computer and segmentation algorithm, in this article, we present a new approach to tumor detection and localization. This is an automatic method, easy and simple to program and requires little simulation time, it is based on unsupervised classification which dont need to reference image, so it is completely automatic. This method is based on two steps; the first is the classification of the MRI image into two classes (class tumor and non tumor class) using the K-means classifier, and the second is the tumor extraction finalization by the thresholding process.
Keywords: Cerebral Tumors, MRI, unsupervised classification, K-means, thresholding
Download Full-Text
ABOUT THE AUTHORS
Sara Sandabad
Laboratory LRGE, ENSET, Mohamed V University Rabat, Morocco.
Achraf Benba
Laboratory LRGE, ENSET, Mohamed V University Rabat, Morocco.
Yassine Sayd Tahri
Laboratory LRGE, ENSET, Mohamed V University Rabat, Morocco.
Ahmed Hammouch
Laboratory LRGE, ENSET, Mohamed V University Rabat, Morocco.
Sara Sandabad
Laboratory LRGE, ENSET, Mohamed V University Rabat, Morocco.
Achraf Benba
Laboratory LRGE, ENSET, Mohamed V University Rabat, Morocco.
Yassine Sayd Tahri
Laboratory LRGE, ENSET, Mohamed V University Rabat, Morocco.
Ahmed Hammouch
Laboratory LRGE, ENSET, Mohamed V University Rabat, Morocco.